[00:02] <maxb> yes
[00:04] <Carl0> We just installed Bzr on OS 10.6 and get an error whenever we try to init a repository: bzr: ERROR: '/Users/carl' is not a working copy
[00:05] <Carl0> I cannot find other users talking about this error when I searched Google. Can anybody here please help us?
[00:08] <maxb> Please give the *precise* command you ran, and run it again with the -Derror flag to see if it gives more information
[00:10] <carl_> I accidentally disconnected. Can anybody help us figure out the cause of the following error when we init a repository: bzr: ERROR: '/Users/username' is not a working copy
[00:11] <maxb> Please give the *precise* command you ran, and run it again with the -Derror flag to see if it gives more information
[00:11] <carl_> We just installed bzr 2.2b3 on OS 10.6
[00:11] <Stavros> maxb: how?
[00:12] <carl_> $ bzr init
[00:13] <carl_> .. UnsupportedFormatError: '/Users/carl' is not a working copy
[00:14] <maxb> Is there already a /Users/carl/.bzr/ directory? or .git ? or .svn ? or .hg ?
[00:14] <carl_> http://pastebin.org/397081
[00:14] <maxb> Stavros: have you installed the bzr-hg plugin?
[00:15] <carl_> maxb, there is only a .bzr.log
[00:15] <maxb> carl_: OK, so bzr-svn is somehow breaking things
[00:16] <carl_> Can we remove bzr-svn?
[00:17] <Stavros> maxb: yes, but it's not working :/
[00:17] <Stavros> i do bzr branch hg+http://url and it says that the scheme is not supported
[00:17] <fullermd> Last I heard bzr-hg didn't support remote protocols.
[00:18] <maxb> carl_: I would just delete it or move it aside. But then, I usually run bzr-svn directly from a branch. I am not familiar with OS X or what bzr installers may be there
[00:18] <Stavros> oh
[00:18] <Stavros> hmm, too bad...
[00:18] <fullermd> (or really much of anything beyond PoC)
[00:18] <maxb> fullermd: there are live bzr-hg imports on launchpad...
[00:18] <maxb> Stavros: lose the hg+
[00:19] <fullermd> Well, it's been years since I acquired the "last I heard", to be sure.  But it's always been a long time since I heard anybody actively working on bzr-hg too.
[00:19] <fullermd> s/always/also/
[00:20] <maxb> It's a lot less polished than bzr-svn, that's for certain, but it's often good enough
[00:21] <fullermd> The last I heard was that it was capable of branch'ing and pull'ing and such from a local hg branch, but not over remote protocol, or able to push back.
[00:35] <carl_> maxb, thanks for suggesting the svn folder removal.
[00:36] <carl_> Now when I try to init a repository I get an error telling me 'No repository present'
[00:36] <carl_> here is the traceback: http://pastebin.org/397096
[01:33] <Meths> What does bzr do on a commit after printing "Committed revision xyz."?  Mine seems to take quite a while to return to the command prompt compared to what it used to.  Noticed with 2.1.x on WinXP.
[08:38] <janisozaur> in svn I can "svn log BASE:HEAD", this shows all the commit logs that happened since last "svn up" - local revno is compared against server's revno. How can I achieve the same in bzr?
[08:39] <janisozaur> I've tried "bzr log -r -1:lp:cybercarto" (as cybercarto is the project I manage) but bzr tells me that it had encountered an internal bug. Should I file it?
[08:40] <janisozaur> *internal error
[08:42] <parthm> janisozaur: internal error is almost always a bug. it would be good to file it.
[08:42] <swathanthran> bzr commit won't look at $VISUAL? or which variable i should set to make it point to an editor i want?
[08:43] <parthm> swathanthran: EDITOR or BZR_EDITOR. `bzr help env-variables`
[08:43] <janisozaur> parthm: do I need to attach only the crash log or should I also include snapshot of local and server state of branches?
[08:44] <parthm> janisozaur: crash log should be enough.
[08:44] <swathanthran> parthm: thanks
[08:44] <parthm> janisozaur: the way i do what you mention is that i keep a mirror of the trunk around
[08:44] <parthm> janisozaur: then from my feature branch foo, i can do "bzr diff --old ../trunk"
[08:45] <parthm> janisozaur: typically i keep all my branches in a shared repo locally so they share history.
[08:46] <janisozaur> parthm: so... there is no way of fetching logs only for missing commits from server?
[08:48] <parthm> janisozaur: you don't really need to connect to the server as all the logs are available locally.
[08:48] <parthm> janisozaur: you could do a `bzr log | less` and that has the entire log.
[08:49] <parthm> janisozaur: if you do `bzr revno lp:foo` you will get the revno of the trunk. you could then do `bzr log $revno..last:1`
[08:50] <janisozaur> parthm: yes, they are stored locally, but what I'm used to, after using some svn, is checking the logs before updating.
[08:53] <parthm> janisozaur: i think the revno + log which i mentioned above should do what you want.
[08:55] <parthm> janisozaur: unless the trunk has moved ahead. i which case i don't know of a good way to accomplish what you mention.
[10:01] <jam> janisozaur: bzr log -r -1..-1:. lp:cybercarto will probably work
[10:01] <jam> sorry
[10:02] <jam> you want to make sure one is the local, and one is the remote
[10:02] <jam> but that you call "bzr log REMOTE -r local..remote"
[10:04] <jam> I haven't found a syntax for the local branch that works as the first part of a range yet, checking
[10:05] <jam> janisozaur: vila points out that "bzr missing --theirs" also works for this
[10:07] <jam> janisozaur: bzr log lp:bzr -r -1::this..-1
[10:10] <jam> its sort of magical syntax
[10:10] <jam> but it does do what you want
[10:10] <jam> missing can certainly be a lot easier, though
[10:23] <ronny_> hi
[10:24] <ronny_> can someone please make at least the recent bzr pip installable?
[10:25] <ronny_> hum, it actually half-works
[10:51] <ngirard> Hi all. I need to know about this while not beeing able to access bzr on the command line right now:
[10:51] <ngirard> when calling
[10:51] <ngirard> bzr export DEST [BRANCH]
[10:52] <ngirard> will bzr create DEST if it doesn't exist ?
[10:56] <ngirard> anyone ?
[11:09] <bialix> hey
[11:11] <mwhudson> ngirard: yes
[11:15] <ngirard> mwhudson, thanks ! I'm currently writing a shell script that i cannot test now ;-)
[11:17] <mwhudson> ngirard: that's not a good place to be!
[11:18] <ngirard> well mwhudson, your answer was the sole thing I needed there !
[13:13] <ngirard> I'm only using bzr do  export some stuff that'll never be modified. How can I get rid of the "you have not informed bzr" message ?
[13:14] <beuno> ngirard, ser a name?
[13:17] <ngirard> aha. I can bzr launchpad-login --no-check. That's fine. Thanks beuno
[14:31] <d1b> morning
[14:31] <d1b> can i use bzr to pull from an svn repo and then push to a git repo?
[14:31] <d1b>  / does bzr understand / work with svn branches?
[14:31] <poolie> d1b, yes, and yes
[14:32] <d1b> poolie: how do i do this? :P
[14:32] <d1b> like i mean easy
[14:33] <poolie> bzr branch svn+ssh://example.com/foo
[14:33] <d1b> erh the svn is https :// :P
[14:33] <poolie> bzr push -d foo git+ssh://example.com/git
[14:33] <d1b> and what if i want to do a local git export
[14:33] <poolie> well, that then
[14:33] <spiv> d1b: I bet your host isn't example.com either :)
[14:33] <d1b> spiv: :P
[14:34] <d1b> actually this isn't going to work either lol because i would need to merge git and git-svn
[14:34] <d1b> mmm
[14:34] <d1b> still history in bzr or git would be better than svn
[14:36] <d1b> erh what if the remote https location has an invalid cert?
[14:37] <spiv> try https+urllib: instead of https:
[14:39] <d1b> nope fails :/
[14:39] <d1b> doesn't like my username / pass
[14:39] <d1b> or something :/ -both are valid
[14:40] <lifeless> have you logged in with svn ?
[14:40] <d1b> git-svn works as does svn
[14:40] <d1b> i assume you guys are using pycurl
[14:40] <lifeless> no
[14:40] <d1b> if so, please add the support to add an external cert
[14:41] <d1b> ok
[14:41] <lifeless> you're talking to svn
[14:41] <d1b> same thing holds :)
[14:41] <lifeless> we use libsvn
[14:41] <d1b> i am talkingto svn :)
[14:41] <d1b> it doesn't like the user / pass and both are correct
[14:41] <d1b> it starts to do the checkout then stops
[14:41] <d1b> and reasks
[14:41] <d1b> "bzr error not a branch"
[14:42] <lifeless> d1b: firstly don't pass the user/pass in the url, for svn if you have logged in already its implicit
[14:42] <d1b> lifeless: im not
[14:42] <lifeless> secondly, perhaps you can run with -Derror and show what happens
[14:42] <d1b> why would i put the username in the url :) ? ok will do
[14:45] <d1b> http://paste.pocoo.org/show/237935/
[14:46] <d1b> perhaps i need to say master?
[14:46] <d1b> in which case this would violate my "easily get my svn branches"
[14:50] <d1b> perhaps it is getting the username wrong ?
[14:50] <d1b> (appending @example.com)
[15:06] <d1b> ok so no ideas?
[15:07] <svaksha> while trying to pull this branch, https://code.launchpad.net/~mailman-coders/mailman/2.1, i get this error: bzr: ERROR: Invalid url supplied to transport: "lp:mailman-2.1": No such project: mailman-2.1
[15:09] <Ursinha> svaksha, wouldn't that be lp:mailman/2.1?
[15:09] <svaksha> that is what i entered
[15:09] <Ursinha> I mean, what command you ran?
[15:10] <svaksha>  sudo bzr branch lp:mailman/2.1
[15:10] <svaksha> which gave the above error
[15:10] <svaksha> so i tried it with -
[15:10] <svaksha> s/above/
[15:11] <svaksha> bzr: ERROR: Unknown branch format: 'Bazaar Branch Format 7 (needs bzr 1.6)\n'
[15:12] <poolie> svaksha, maybe you need a newer bzr?
[15:12] <d1b> upgrade your bzr!
[15:13] <lifeless> d1b: if you have a branch called master, under foobar, then yes you need to specify it
[15:14] <svaksha> poolie: i re-installed bzr a couple of days ago. *sigh*
[15:14] <lifeless> d1b: the 'svn-import' command will however pull all the branches
[15:14] <poolie> svaksha, what platform are you on?
[15:14] <svaksha> poolie: hardy
[15:14] <d1b> lifeless: how how does that work?
[15:14]  * svaksha had some problems with bzr so i purged the old installation and reinstalled
[15:14] <d1b> do i just add that to bzr branch?
[15:14] <poolie> svaksha, use https://edge.launchpad.net/~bzr/+archive/ppa
[15:15] <d1b> wow my brain is on stupid today
[15:15] <svaksha> poolie: thanks. will get back to you on it
[15:16] <d1b> ah right got it
[15:16] <d1b> however, i put /trunk before - which does exist and that didn't work either
[15:23] <d1b> bah turns out bzr-svn doesn't even do what i want. oh well. - according to the site it doesn't truncate branch svn branch copies
[15:24]  * maxb wonders what "truncate svn branch copies" means
[15:25] <d1b> maxb: i assume it means doesn't maintain the history of a file when copied toa branch
[15:25] <d1b> which is exactly what git is getting confused with - i stupidily didn't make this change from git and now it is lots of fun!
[15:26] <d1b> it works fine but i wanted to maintain / apply previous file history to the 'new' (copied) branch
[15:26] <svaksha> poolie: on the development server i'm only allowed to install packed versions of software, so a ppa is not workable. is there another way to pull this Mailman revision?
[15:27] <poolie> svaksha, you realise the things in there are packaged?
[15:28] <svaksha> poolie: yes, i do.
[16:40] <poolie> svaksha, branch it onto a different machine using an uptodate client i guess
[16:52]  * maxb looks at .bzr.log and recoils at the amount of spurious errors bzr-svn is logging
[16:52] <svaksha> poolie: i was suggested another method -- edited sources.list to add the hardy ppa, sudo upgrade +update allowed me to pull the branch
[16:52] <poolie> uh yeah
[16:52] <poolie> that's what i was suggesting you do
[16:52] <svaksha> :) thanks
[17:59] <sven_sandberg> hi! i have encountered what looks like wrong conflict markers when merging one tree to another. it looks similar to this: http://pastebin.com/gzxtEt2X
[18:00] <sven_sandberg> note that after the merge, line 3 has moved from outside the conflict to TREE. so if you think you'd want to take the OTHER part of the conflict, you'd miss line 3.
[18:01] <sven_sandberg> i can't reproduce it in this simple form, only in my local mysql tree. so it's probably history-dependent
[18:02] <maxb> That sounds familiar
[18:02] <sven_sandberg> i know these things can be counter-intuitive sometimes. is this likely to be a bug?
[18:02] <maxb> Yes, that's definitely happened to me too. Unfortunately I've done no better than you at reproducing it
[18:03] <sven_sandberg> is there a good place where i can upload the source tree so that others can reproduce it?
[18:03] <sven_sandberg> maxb, or maybe you can already reproduce it in a public tree?
[18:03] <maxb> my broken tree is private code
[18:04] <maxb> Maybe you could just put the branches on launchpad and file a bug including instructions of what to merge where to reproduce?
[18:04] <sven_sandberg> sure. how should i put it on launchpad? is there a 'scratch area' for things like this?
[18:11] <maxb> sven_sandberg: You can push things to ~your-name/+junk/whatever
[18:11] <sven_sandberg> maxb, thanks!
[18:12] <maxb> Though if it's a branch of an established project, you're better off pushing it at lp:~you/project/tmp-something, so that the auto-stacking takes effect and you don't need to spend network traffic pushing the common history
[18:19] <sven_sandberg> thank you. i'll see what works easiest for me